f\.

Proceedings: Second International Conference on Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics,
March 11-15, 1991, St. Louis, Missouri, Paper No. 4.10
Retaining Walls; Computation of Seismically Induced Deformations
Sreenivas Alampalli
Engineering Research Specialist I, Engineering R & D Bureau,
NYSDOT, Albany, NY 12232, USA
Ahmed-W. Elgamal
Assistant Professor, Department of Civil Engineering, Rensselaer
Polytechnic Institute, Troy, NY 12180, USA
SYNOPSIS: A two dimensional (2D) dynamic wall-soil computational model is proposed. The model accounts
for wall and soil resonance, nonlinear wall-backfill soil interaction, simultaneous wall base sliding and
rotation, nonlinear soil properties and possible pore pressure buildup. A bending beam with a base
yielding rotational spring and a base translational slide element represents the wall. A 2D shear beam
represents the soil system. The wall supporting soil through a Yinkler type
no-tension spring system. An elasto-plast1c path-dependent hysteret1c model accounts for nonl1near so1l
behavior and possible pore pressure buildup. The seismic response of a 15m high cantilever wall is studied
in detail. Yall translational and rotational failure mechanisms are discussed. The computed results
indicate the importance of including seismically induced moments in dynamic wall stability ev_aluation. It
is also found that retaining walls with loose saturated backfill soils may accumulate excess1ve permanent
displacements well beyond the strong shaking phase of an earthquake.
INTRODUCTION
A number of analytical models which investigate
the dynamic behavior of retaining wall systems
have been proposed in the past. Taj imi (1973) used
an elastic two dimensional wave propagation theory
to investigate earth pressures on a basement wall
assumed to undergo periodic vibrations of
horizontal translation and rotation. Scott ( 1970)
proposed a one dimensional elastic shear beam to
model backfill soil connected to a wall
by a system of Yinkler springs. Arias et al. t 1981)
developed a 2D shear model to analyze fixed rigid
wall response.
In addition to modeling the vibration of the
wall-soil system, some nonlinear soil-structure
interaction aspects may have a major influence on
potential plastic deformations. Yall-backfill
soil as well as wall-base interaction control the
magnitude of dynamic earth pressure and resulting
wall deformation. Prakash (1981) outlines a
method for calculating permanent sliding
displacement of retaining walls. In this method,
wall and soil are modeled as a single degree of
freedom mass supported by a nonlinear yielding
base spring. Prakash et al. (1981), Nadim and
Yhitman (1984), and, Siddarthan et al. (1990) show
that the rotational deformation of the wall
structure may be quite significant in some cases
and should be accounted for in analysis
procedures. Nadim and Yhitman (1983), and,
Siddarthan et al. (1989) investigated retaining
wall response using a nonlinear plane strain
finite element (FE) analysis incorporating slip
elements along the active failure wedge boundaries
and the wall base-soil interface. Prevost (1985)
and Siller et al. (1987) investigated the dynamic
response of a retaining wall using an
635
elasto-plastic 2D FE model. Importance of
nonlinear soil properties and amplification of
backfill motion in dynamic analysis of wall soil
systems is emphasized in the above mentioned FE
studies.
In this paper a model which accounts for wall and
soil resonant dynamic response is proposed. The
retaining wall is represented by a bending beam
with a yielding base rotational spring and a
translational slide element. Backfill soil is
represented by a 2D shear beam. The wall and
supported soil interact through a Yinkler type
nonlinear no-tension spring system. A formulation
is developed in which the free vibration mode
shapes of the wall and the soil is employed. An
elasto-plastic constitutive relation is
incorporated to allow for nonlinear soil
properties. Soil strength and stiffness
degradation with pore pressure increase is also
allowed. The seismically induced translational
and rotational failures of a 15 m high cantilever
retaining wall are studied using this model.
FORMULATION OF GENERAL MODEL
A wall model is assumed to interact under dynamic
loading with a soil model through a system of
Yinkler-type springs. Two dimensional in-plane
vibration conditions are assumed. Response
features incorporated in this model are discussed
below followed by details of the formulation.
Features Of Proposed Dynamic Model
In this section, the features incorporated in the
proposed dynamic wall-soil model (see Fig. 1) are
presented. Some of these features are included in
currently available models and some are unique to
this model. These features are:
a) Simultaneous wall translation and rotation: A
translational slide element is available at the
retaining wall base. Under dynamic loading
conditions, the sliding yield force of this
element is:
F = Active Earth Pressure Force x
y
( F.S. against sliding-1.0) (1)
where F. S. is factor of safety. F y may be further
adjusted to include the effects of earthquake
vertical shaking, and of increased static earth
pressure and reduced base friction coefficient due
to pore pressure buildup.
An elastic perfectly plastic
also available at the base.
chosen to account for wall
under dynamic loading:
M = Overturning Moment x
y
rotational spring is
The yield moment is
rotational failures
(F.S. against overturning-1.0) (2)
Alternate expressions for MY may be defined if wall
rotation occurs about points other than the wall
toe. MY may be also chosen to represent the
resistance to bending of a plastic hinge which may
develop at the wall stem-base juncture.
Modifications to account for effects of
pore-pressure buildup are also possible.
b) No-Tension Vall-Soil Interface: A nonlinear
Vinkler type spring system is chosen as a wall-soil
interface. No tension properties are intended to
represent the dynamic component of wall-soil
interaction. These nonlinear Vinkler springs mar
be adapted to fill any gaps (partially or fully)
created during transient seismic interaction.
Such gaps may occur due to wall sliding, wall
rotation or due to difference in inertia and
stiffness of wall and soil. The nonlinear Vinkler
springs may also be used to exert additional
stresses on the wall which may arise from backfill
densification.
c) Flexible Vall Model: An Euler bending beam
represents the wall. In many practical cases one
or two mode shapes of this beam (1 or 2 degrees of
freedom) will provide sufficient accuracy in
defining the dynamic wall response. This wall
model provides a realistic boundary for
wall-backfill dynamic interaction. Gravity as
well as flexible walls may be modeled. In
addition, a base mass may be included to simulate
the wall foundation.
d) Soil Model Vith Ground Motion Amplification: A
simple soil model is proposed. Only lateral shear
vibration is included (2D shear beam). In most
cases, few degrees of freedom (10 or less) will
represent soil response with sufficient accuracy.
As mentioned earlier, no-tension Vinkler springs
allow wall-backfill soil interaction. Other
backfill boundaries (base and far end) are modeled
by appropriate spring-dashpot mechanisms. Note
that the soil base boundary may be ali&ned with the
available geological profile (Fig. 1). Boundary
springs and dashpots can be chosen so as to account
for far-field compliance and radiation damping
effects. Elasto-plastic nonlinear hysteretic
soil properties may be included in this model.
636
Reteining
Well
Retaining Well
k
Soil
H,
u g(t)
Fig. 1.
Wall-Soil Configuration And Corresponding
Computational Model.
I
l
u g(t)
In summary, the following aspects of wall-soil
response are accounted for: wall flexibility,
simultaneous wall translation and rotation,
no-tension wall-soil interface, actual (2D) soil
geometry, amplification of dynamic/earthquake
input motions, radiation damping effects, and,
soil elasto-plastic nonlinear response. A small
matrix equation (15x15 or less) will in general
incorporate the above features and provide
solutions of sufficient accuracy.
1990 and
The retaining wall is represented by a one
dimensional Euler bending beam with a fixed base.
Dynamic soil pressure on the wall is represented by
continuous Vinkler type springs alongside the wall
as shown in Fig. 1. Natural frequencies and mode
shapes i(z), i = 1,2, ... ro of this system can be
easily obtained. The bending beam described above
will be supported on a stick-slip frictional base
of mass mbs and a yielding torsional spring KT
(Fig. 1). This base support system (degrees of
freedom ubs in translation and ()in rotation) will
be used to allow for wall sliding and overturning.
Lateral vibration of the backfill is represented
by a two dimensional soil domain. Natural
frequencies and mode shapes ~ (x,z), i=1 ,2 ,3, ... ro
can be derived in closed form for this soil system.
The above discussed wall and soil models are used
to represent the vibrational response of combined
wall-soil systems. As mentioned earlier, the wall
base will be allowed one translational and one
rotational degree of freedom. In translation, the
wall will be allowed to slide away from the
backfill once the dynamic lateral forces exceed
the base frictional force as dictated by the static
factor of safety against sliding. An additional
elastic-perfectly-plastic base rotational
will also allow rotation (away from the backfill)
if base moment exceeds the available resistance as
dictated by the static factor of safety against
overturning.
In the present analysis a simplified version of the
above described general wall-soil model is
employed (see Fig. 2). In this simplified model,
the following assumptions are made: 1) Soil far
end is fixed, 2) Soil base is fixed, 3) Vall height
is equal to soii domain height.
The wall will interact with soil through a
no-tension Vinkler type springs (k) as shown in
Fig. 3. Under dynamic excitation, these springs
will sustain compressive forces only. It is noted
that the force-displacement relation of the
springs (k) will always fill any tensile gaps and
maintain continued interaction between the wall
and the backfill. Considering interaction forces
and ground motion excitation, the equations of
motion of the wall-soil system can be obtained.
After multiplying these equations by a variation
(v(x,z) for soil equation and vb(z) for wall
equation), integrating over the respective
spatial domains, performing partial integration
and substituting boundary conditions, the
following equations are obtained:
6 [pu,ttv+Gu,zv'z+Eu,xv'x+pvug'tt] dfl
H
+ b k [u(x=O) -ub -ubs -z 0] v(x=O) dz = 0
(n =Soil Spatial Domain x,z) (3)
H
b [m ub'tt vb + E'I ub'zz vb'zz + m ug'tt vb
(4)
where u is the relative displacement of soil at any
point (x,z), G is the shear modulus of soil, E is
the Youngs modulus of soil, pis the mass density of
soil, L is length of soil domain, H is height of
soil domain/wall, ub is relative displacement of
the wall at any point z, E' is modulus of elasticity
of the wall material, I is wall moment of inertia, m
is wall mass per unit length, ug is ground
displacement, ubs is relative wall base
displacement, 8 is wall base rotation (radians),
and, tis time.
Employing the mode shapes i' i=1,2, .. n for the
beam j=1,2, .. m for the soil domain to
J
637
Well-Soil lnterfece Springs (k)
Soil System
(2D Sheer Beam)
H

Fig. 2. Simplified Wall-Soil Computational Model.
Fig. 3.
Spring Force
Wall-Backfill Soil Interaction Nonlinear Spring
Formulation.
represent the solution, a matrix equation is
?btained. Proportional viscous damping can be
1ntroduced as a weighted combination of the
inertial and stiffness terms. Newmark's
predictor-multi-corrector implicit scheme with or
without user specified numerical damping is
empl?yed to o_btain a step-by-step solution of the
matr1x equat1on. Note that, due to the various
incorporated nonlinearities, iterations will in
general be performed at each time step so as to
achieve a specified convergence tolerance.
Elasto-Plastic Constitutive Model (Elgamal and
Alampalli, 1990)
A soil elasto-plastic path-dependent hysteretic
model is in the 2D computational
model descr1bed above. The flow or incremental
theory of plasticity is used. The model is capable
of the behavior (Massing,
1926) :xh1b1ted by so1l under cyclic loading. A
harden1ng rule proposed by Mroz and Zienkiewicz
(1984) is This hardening rule
leads to an 1nf1n1te y1eld surface formulation in
which evolution of stress state is piecewise
linear between computation steps (see Fig. 4). A
simple pore pressure generation logic is also
included in order to account for degradation of
stiffness and strength in loose saturated sandy
soils (see Fig. 5).
N
E

"'
.n

n

OJ
.c
'f;
Fetlure Surfe.ce /
A Current Stress Stele
B Current Vteld Surf ece (f y)
c. current Outer Surfece (1
0
)
o stored Memory Surfl!lces
E canceptue.l Represenll!lt1on or interpole.tton surfeces
Fig. 4.
Conceptual Configuration Of Hardening
Process In General Stress Space.
3.00'
2.50 t
2.00 I
1.50 +
1.00 +
0.50
0.00
I
-0.50 +
-1.00
-1.50
ia)
SEISMIC RETAINING VALL RESPONSE
In this section, the proposed simplified model is
used to investigate the seismic response of a
reinforced concrete cantilever retaining wall and
the associated seismically induced base sliding as
well as base rotational failure mechanisms. The
cantilever retaining wall to be modeled is of 15 m
height and 1. 2 m thickness supporting a 150 m long
soil backfill {Fig. 6). Various properties of the
system are: E' = 2. 2 x 10
9
Kgjm
2
, m = 240 Kg sec
2
/m
4
,
k = 5 x 10
5
kgjm
2
, E = 3.12 x 10
7
Kgjm
2
, Poisson's
ratio of soil 0.3, p 216 Kg sec
2
/m\
proportional viscous damping= 37., soil internal
friction angle = 36 degrees, and KT = 4.5 x 10
9
Kg m/radian. Six beam mode shapes and 10 soil mode
shapes are employed.
Retaining walls in general are designed for a
factor of safety (F.S.) of at least 1.5 against
possible base sliding or overturning. Only the
resistance in excess of the design static loads
will be available to resist additional forces on
the wall when subjected to dynamic/seismic
excitation. The values given by Eqs. 1 and 2 are
used as yield limits for the wall base
translational slide force and rotational spring
moment respectively. Static earth pressure is
evaluated using the Rankine theory. Scaled
El Centro 1940 SOOE earthquake acceleration record
is used to represent ground {base) excitation in
the present analysis. The maximum lateral
acceleration of this scaled earthquake is 0. 60 g.
1) Vall-Dry Backfill Soil
Properties)
(Linear Soil
Time histories of base excitation, absolute
acceleration at soil surface {adjacent to the
wall), permanent displacement accumulated at wall
top, and, stress at 3m above the wall base are
shown in Fig. 7 (F. S. = 1. 5 against both base
rotation and base translation). Based on these
results, the following conclusions can be deduced:
a) Input excitation is significantly amplified in
backfill soil {about 2.5 times), b) A residual
pressure is maintained on the wall at the end of
shaking. Note that in this model, interface
springs always close any tensile gaps developed at
the wall-soil interface, c) Most of the
accumulated wall permanent deformation coincides
with the strong shaking phase of the earthquake
(during the first 3 seconds). This deformation is
predominantly due to the base rotation since, in
this case, no base translation occurred
(Alampalli, 1990).
+ 2) Vall-Dry Backfill Soil (Nonlinear Soil
1 Properties)
t
-2.00 'i-.,.::::=:;::::::..--+---+-----i-------<---+--+--+
The same computational model is used along with the
soil plasticity model. In this plasticity model,
the hyperbolic stress-strain curve fitting
parameter is set to the value of 0. 0013 (Alampalli,
-0.5 -0.4 -0.3 -02 -0.10 0.00 0.10 020 0.30 0.40 0.50
Fig. 5.
Shear strain ( x 10 -2)
Material Stress-Strain Behavior Under Cyclic
Loading (Strain Controlled).
638
1990). Viscous damping in soil is not employed.
Damping in soil is consequently only hysteretic.
Time histories of soil surface acceleration and
wall top relative displacement are shown in Fig. 8.
The following observations can be made by
comparing these computed results {Fig. 8) with
those of Fig. 7: a) amplification of input
0
0.60
X 0.50
N
0.40
u
0.30 QJ
Vl
......
0.20
E
0.10
c:
0
0.00
~ -0.10
...
,!!!. -0.20
~ -0.30
:i. -0.40
0.00
(a)
......
I
C>
000
......
X -1.00
5-2.00
~ -3.00
Q)
~ -4.00
u
:: -5.00
c.
. ~ -6.00
Cl
-7.00
~ -B.OO
0.00
(c)
Fig.
Retaining
Wall
\ ~ l ~ m ~
Soil
Fig. 6.
Cantilever Retaining Wall-Soil System Employed In Numerical Simulation.
0
0.6C
......
X
1.20
N
O.BO u
QJ
Vl
...... 0.40
E
0.00
c:
. ~ -0.40
+-'
"' ... -0.80
QJ
a:; -120
u
u
c:t: -1.50
-0.46
0.30 0.60 0.90 120 1.50 LBO 2.10 2.40 0.00 0.30 0.60 0.90 120 1.50 LBO
Time (seconds) X 10 Time (seconds) X 10
Base Excitation (b) Soi 1 Surface Acceleration
o.os
2.40
...,
0
2.10
X
l.BO
N
1.50
E
......
0')
1.20
""
QJ
0.90
...
:::l
0.50
Vl
Vl
QJ
0.30
...
c..
-B.32 000
0.30 0.60 0.90 120 L50 1.80 2.10 2.40 0.30 0.60 0.90 1.20 1.50 LBO
Time (seconds) X 10 Time (seconds) X 10
2.10 2.40
2.10 2.40
Wall Top Relative Displacement (d) Wall Pressure At 3m Above The Base
7. Numerical Simulation Of Wall-Soil Earthquake Response ( L i near So i 1 Properties).
639
!.56
-1.89
2.61
DOC
0
X 0.60
N
~ 0.40
"'
E"" 0.20
<= 0.00
0
~ -0.20
s...
~ -0.40
Q)
u
;;:}_ -0.60
0.30 0.60 0.90 1.20 1.50 1.80 2.10 2.40
Time (seconds) x 10
(a) Soil Surface Acceleration
......
I
0
......
><
E
...,
<=
Q)
E
Q)
u
"'
c.
"' ~
Cl
Q)
""
0.00
-0.30
-0.60
-0.90
-120
-1.50
-1.80
0.30 0.60 0.90 1.20 1.50 1.80 2.10 2.40
Time (seconds) x 10
(b) Wall Top Relative Displacement
Fig. 8. Numerical Simulation Of Wall-Soil Earthquake Response (Nonlinear Soil Properties).
excitation is noticeably decreased when nonlinear
soil properties are used, b) permanent wall base
rotation is also smaller when nonlinear backfill
properties are specified.
3) Vall-8aturated Backfill Soil
Seismic response of the above mentioned retaining
wall-soil system is computed after modifying the
computational model to account for possible
increase in pore pressure. The pore pressure model
parameters employed are based on loose Ottawa sand
properties (Alampalli, 1990). For simplicity a
ground water table is assumed to coincide with
ground surface on the backfill side only. Only
base translational yielding is allowed and the
initial yield value Fy(O) is chosen based on a
static factor of safety of 1.5 against possible
translation. The initial wall-foundation soil
friction angle cb(O) is assumed to be equal to the
angle of internal friction.
The base translational yield force Fy(t) changes
at every time step depending on static active earth
pressure increase and decrease in wall
base-foundation soil friction angle (cb) with
increase in pore water pressure. It will be chosen
to change in the following manner:
where,
R(t) =tan cb(t) jtan 6b(O)
F i (t) =Increase in static earth pressure force
on the wall at time t,
tan cb (t) = (1-K(t)) / (2 .;-Ir(t}), and,
K(t) = Coefficient of active earth pressure at
timet.
640
In the above equation, K is assumed to vary with the
change in excess pore pressure ratio ( ru, 0 ~ ru ~ 1
in the following simple fashion:
K(t) = (1-ru (t)) K(O) +ru (t).
In computing F (t), the excess pore pressure ratio
y
(r ) developed in the soil near the wall base is
u
employed. In computing Fi(t), excess pore
pressures along the wall-soil interface are
employed. It is noted that the above relations for
F (t) and K(t) are not intended to represent any
y
particular field situation. Relations which
govern the change in active wall pressure, and
buildup of excess pore pressures below a retaining
wall base, may be quite different from those
adopted herein ( eg., see Elgamal et al., 1989).
The obtained time histories of excess pore
pressure ratio near the wall base (at the wall-soil
interface), wall base displacement, base yield
force, base force, wall pressure at 3m above the
wall base, along with shear stress vs shear strain
(near wall base) are presented in Fig. 9. From
these results (Fig. 9), the following is observed:
a) The excess pore pressure ratio (ru) near the
wall base (at the wall-soil interface) reaches a
very high value of 0. 83 by the end of first the 15
seconds of earthquake excitation. A value of 0.541
is attained in the soil close to the wall top
(Alampalli, 1990). It is noted here, that in an
actual situation, wall translation will lead to
the development of very large shear strains in the
soil system. These strains (not accounted for in
this model) will undoubtedly influence
pore-pressure buildup along the wall height.
b) The retaining wall moves as much as 1. 20m at the
end of the first 14 seconds. Base accumulated
translation continues throughout the duration of
shaking however. In fact after about 14 seconds,
all residual base translational resistance is gone
due to pore-pressure build-up and an additional
I
8.00
0
7.00
X
0
6.00
....,
"'
5.00
a:
<lJ 4.00
,_
:::t
3.00 Vl
Vl
<lJ
,_
2.00
0..
<lJ
1.00 ,_
0
0..
000
(a)
7.00
..,.
6.00
X 5.00
E
';;:, 4.00
"'
3.00
<lJ
u
2.00
u...
;:: 1.00
w
;: 0.00
0.30 0.60 0.90 120 1.50 1.80 2.10 2.40
Time (seconds) X 10
Excess Pore Pressure ratio Near The Wall
0.30 0.60 0.90 120 1.50 1.80 2.10 2.40
Time (seconds) x 10
(c) Wall Base Yield Force
2.70 +-r--+----t----t----t---+---+---+---t---+
2.40
(V) 2.10
0
1.80
X

N
1.20
0>
.:::_ 0.90
0.60
:::>
0.30
<lJ

0.30 0.60 0.90 120 1.50 1.80 2.10 2.40
Time (seconds) x 10
(e) Wall Pressure At 3m Above Wall Base
8.32
O.OC
Base
7.31
oon
272
O.DC
0
O.OC
X
-0.30
E
-0.60
....,
-0.90
"'
<lJ
E
-120
<lJ
u
"'
-1.50
n.
"'
-1.80
0
-2.10
-2.40
<lJ
-2.62
a:
0.30 0.60 0.90 1.20 1.50 1.80 2.10 2.40
Time (seconds) X 10
(b) Wall Base Relative Displacement
2.9S
2.50
2.00
..,.
1.50
0
1.00
X
0.50
fmfflvr--
E
...__
000
0>
n r
I
I
"' -0.50
<lJ
-1.00 u
,_
0
..... -1.50
'F--'--+----+---+---+---+---+---+--+-----+ -1.57
..,.
0
X
N
E
...__
Ol
"'
Vl
Vl
<lJ
,_
....,
V)
,_
"'
<lJ
..c:.
V)
0.30 0.60 0.90 1.20 1.50 1.80 2.10 2.40
Time (seconds) x 10
(d) Wall Base Force
1.00
0.80
0.60
0.40
0.20
0.00
-0.20
-0.40
-0.60
-0.80
(f)
-3.00 -2.00 -1.00 0.00 1.00 2.00 3.00 4.00
Shear Strain x Io-3
Shear Stress vs. Shear Strain Near Wall Base
Fig. 9. Numerical Simulation Of Wall-Soil Earthquake Response (Saturated Loose Soil And Nonlinear Soil
Properties).
641
base translation of 1.42 m occurs. This indicates
that even if the wall does not move considerably
during the initial strong shaking, very large
displacements may occur during the tail end of an
earthquake during an immediate after-shock.
High developed in the
initial strong-motion phase (0-7 seconds)
significantly reduce available base resistance
and lead to large accumulated displacements
thereafter.
c) After about 12 seconds, the wall slides freely
and is unable to sustain any imposed dynamic
pressure.
SUMMARY AND CONCLUSIONS
A two dimensional (2D) dynamic wall-soil
computational model is proposed. The model
accounts for wall and soil resonance, nonlinear
wall-backfill soil interaction, simultaneous wall
base sliding and rotation, nonlinear soil
properties, and possible pore pressure buildup.
An elasto-plastic path-dependent hysteretic model
accounts for nonlinear soil behavior and possible
pore pressure buildup. The seismic response of a
15m high cantilever wall is studied to investigate
seismically induced wall translational and
rotational failure mechanisms. Results indicate
that; 1) retaining walls are vulnerable to
seismically induced rotational failure given an
equal static factor of safety against possible
translation and rotation, 2) input ground motion
may be significantly amplified within the backfill
particularly for tall walls, and, 3) for walls with
saturated loose backfill, large amounts of wall
translation may accumulate after the strong
shaking phase of an earthquake due to the
compromise of base resistance (caused by developed
excess pore pressures) .
It is finally noted that accurate computational
simulation of an actual field condition where high
excess pore pressure may develop is a task of
profound complexity. Considerable research is
still necessary in the areas of large strain
saturated soil behavior and soil-structure
interaction.
ACKNOVLEDGEMENTS
This research is
No. 873011 . The
acknowledged.
REFERENCES
supported
support
by
is
NCEER Grant
gratefully
Alampalli, S. "Earthquake Response Of Retaining
Valls; Full Scale Testing And Computational
Modeling", Ph. D. Thesis, Department Of Civil
Engineering, Rensselaer Polytechnic Institute,
Troy, NY, 1990.
Arias, A., Sanchez-Sesma, F.J., and
Ovando-Shelley, E. "A Simplified Elastic Model For
Seismic Analysis Of Earth-Retaining Structures
Vith Limited Displacements", Proc., International
Conference On Recent Advances In Geotechnical
Earthquake Engineering k Soil Dynamics, Rolla,
Missouri, 1981.
Elgamal, A-V., Dobry, R. and Adalier, K. " Effect
Of Clay Layers On Liquefaction Of Sand Deposits
Using Small Scale Models", Second U.S. Japan
Vorkshop On Liquefaction, Large Ground
642
Deformations And Their Effects On Lifeline
Facilities, Buffalo, NY, 1989.
Elgamal, A-V., and, Alampalli, S.
Response Of Retaining Valls Including
Soil Backfill", submitted for
publication, 1990.
"Dynamic
Supported
journal
Massing, G. "Eigenspannungen And Verfestigung
Beim Messing", Proc. , Second International
Congress Of Applied Mechanics, 1926.
Mroz, Z., and Zienkiewicz, D.C., "Uniform
Formulation Of Constitutive Equations For Clays
And Sands", Mechanics Of Engineering Materials,
Desai, C .S., and Gallaghar, R.H., eds., John Vi ley
& Sons Ltd., 1984.
Nadim, F., and Vhitman,R. V. "Seismically Induced
Movement Of Retaining Valls", Journal Of
Geotechnical Engineering, ASCE, Vol. 109, No.7,
1983.
Nadim, F., and Vhitman, R. V. "Coupled Sliding And
Tilting Of Gravity Retaining Valls During
Earthquakes", Proc., Eighth Vorld Conference On
Earthquake Engineering, San Francisco,
California, 1984.
Prakash, S. "Soil Dynamics", McGraw-Hill Book
Company, New York, 1981.
Prakash, S. , Puri, V. K. , and Khandoker, J . U.
"Displacement Analysis Of Rigid Retaining Valls In
Rocking", Proc., International Conference On
Recent Advances In Geotechnical Earthquake
Engineering & Soil Dynamics, Rolla, Missouri,
1981.
Prevost, J .H. "A Simple Plasticity Theory For
Frictional Cohesionless Soils", Journal Of Soil
Dynamics And Earthquake Engineering, Vol. 4,
No.1, 1985.
Scott, R.F. " Earthquake-Induced Earth Pressures
On Retaining Valls", Proc., Fifth Vorld Conference
On Earthquake Engineering, Rome, 1973.
Siddarthan, R., Norris, G.M., and Maragakis, E.
"Deformation Response Of Rigid Retaining Valls To
Seismic Excitation", Proc., Fourth International
Conference On Soil Dynamics And Earthquake
Engineering, Me xi co, 1989.
Siddarthan, R., Ara, S. and Anderson, J.G.
"Seismic displacements Of Rigid Retaining Valls",
Proc., Fourth U.S. National Conference On
Earthquake Engineering, Palm Springs, California,
Vol. 3, pp. 673-Q82, 1990.
Siller, T.J., Christiano, P.P., and Bielak, J. "On
The Dynamic Behavior Of Tied-Back Retaining
Valls", Developments In Geotechnical Engineering,
Vol. 45, Structures And Stochastic Methods, Third
International Conference On Soil Dynamics And
Earthquake Engineering, 1987.
Tajimi, H. "Dynamic Earth Pressures On Basement
Vall", Proc., Fifth Vorld Conference On Earthquake
Engineering, Rome, 1973.